1. Field of the Invention
The present invention relates to a method for controlling a network comprised of enormous nodes, and a node and a control program to perform said method. More specifically, the present invention relates to a method for controlling roles of the nodes in the network dynamically and distributed autonomously, and a node and a control program to perform said method, in a system to which service is provided by a large number of nodes operating.
2. Description of the Related Art
A sensor network is illustrated as an example of a system to which service is provided by a large number of devices operating. In the sensor network, the sensor device which measures environment not only is used for a single device, but also is used for applications to capture environment information of a wide area with dispersively locating a large number of devices in the environment. On this occasion, since a large number of the devices do not necessarily operate all at the same time, the control structure that only the sufficient number of the devices covering a measurement area operate and the structure to transfer information to an appropriate node in order to collect the information measured by such devices to a data processing server and so on are necessary.
In fact, in this field, the structure to control a large number of devices is actively studied. However, not only the case of comprising of low function and small size devices such as the sensor devices, but also the case of comprising of high-performance devices such as PCs in a P2P network may be considered.
In a conventional computer network, it is previously decided how each node should play a role. Under such limited condition, as the system can suitably operate, various control techniques are devised and implemented. For example, in the case of the sensor network, there is a node called a sink node in a large number of sensor devices (below, it is called a node merely). The Sink node has a gateway role to collect information from the nodes in the network and to hand it to a server in another network (e.g. Internet). When a large number of nodes are sprinkled in the environment, it is necessary to send the information which these nodes maintain to the Sink node efficiently, and then various techniques are devised (For example, “Directed Diffusion for Wireless Sensor Networking,” C. Intanagonwiwat, R. Govindan, D. Estrin, J. Heidemann, and F. Silva, IEEE/ACM Transactions on Networking, VOL. 11, NO. 1, pp 2, 2003 and “GRAdient Broadcast: A Robust Data Delivery Protocol for Large Scale Sensor Networks,” F. Ye, G. Zhong, S. Luand L. Zhang, ACM Wireless Networks, Volume 11, Number 3, pp. 285-298(14), 2005).
Also, as the slightly different technique from this, S. Ratnasamy et al (“Data-Centric Storage in Sensornets with GHT, A Geographic Hash Table,” S. Ratnasamy, B. Karp, S. Shenker, D. Estrin, R. Govindan, L. Yin and F. Yu, ACM Mobile Networks and Applications, Volume 8, Issue 4, pp. 1-16, 2003) changes a maintaining information with depending on a geographical area. They consider a global area that a large number of nodes are sprinkled. When the information obtained at the node of some place is the information of the classification called “A”, the node sends the information to a local area “a” (a plot in the global area). The mapping rule of “a” and “A” is previously determined and the sensor node knows it. Since there are a large number of nodes in the local area “a”, one of them maintains the information. Next, they consider taking the information from the network. Here, if the information “A” is required, a requester, who may be outside the network, transmits a request packet to this sensor network. The sensor nodes converts “A” into the local area “a”, and then transfers the request packet to the local area “a”. And, the node maintaining an appropriate information replies to the request and returns necessary information.